Active Room Treatment (ART) by Dirac

[Mort]

I wonder if ART will limit the support capability offered by your Genelec monitor(s) if it has to work with its GLM-tamed bass response, rather than the potential offered by the much higher SPL “natural” response

I was surprised the Genelec had no such issue. They were eager to help me try this out.

Also curious about the double phase correction, and how having two different measurement microphones factors into it, even if you were able to place them in exactly at the same point in space.

Sound be only a single phase correction. We reset the Genelec phase in step 6.

haven’t tried GLM with ART, so I’m curious as to your results. Might also be worth it for you to try just ART alone.

That's probably not what I'm going to do as I expect the extended phase linearity of glm is more important to me than the bass management ART offers. But will see... That's why I'm testing it.

 

[Wg1]

I was surprised the Genelec had no such issue. They were eager to help me try this out.

Ah by “support” I meant in the sense of ART being able to use the uncorrected bass for whatever it does with how it implements MIMO. Like, maybe uncorrected FR gives ART more to work with, if that makes sense?

It is cool seeing Genelec support this use case and home audio in general.

 

[Mort]

Ah by “support” I meant in the sense of ART being able to use the uncorrected bass for whatever it does with how it implements MIMO. Like, maybe uncorrected FR gives ART more to work with, if that makes sense?

It is cool seeing Genelec support this use case and home audio in general.

Ah. Right. I see what you were saying. Yeah that's a good point. There will be less to work with potentially. I could eliminate any bass DSP in GLM before switching over. That's an interesting idea.

Genelec shows you exactly what they modified.

 

[Defi]

I'm getting very close to going for ART. I have a couple of questions: Is this decay reduction in lower frequencies not possible through other manual DSP, like acourate? Will having 2 subs be enough drivers to perform such cancellations? My instinct is that the more speakers you have that can go down low the better it will be able to adjust for things like that. I am confident I can get an acceptable frequency response from MSO, but I am definitely interested in addressing decay.

 

[Mort]

I have Art up and running on Genelecs that had GLM applied first and stored. It's only a very first run and I didn't measure anything with REW yet. Just an accomplishment to get them up and running on the same system. I can go back and forth in GLM and toggle the output on a Wiim streamer for music. I have to dig into Art further.

 

[dlaloum]

I'm getting very close to going for ART. I have a couple of questions: Is this decay reduction in lower frequencies not possible through other manual DSP, like acourate? Will having 2 subs be enough drivers to perform such cancellations? My instinct is that the more speakers you have that can go down low the better it will be able to adjust for things like that. I am confident I can get an acceptable frequency response from MSO, but I am definitely interested in addressing decay.

People are reporting very positive results even from basic 2.0 traditional stereo setups....

With a 5.1.4 setup vs 5.2.4 - the addition of the extra sub made a very very small difference in decay - no substantive subjective difference in performance - but I believe it may have raised my max SPL headroom (a different dimension in terms of parameters)

However keep in mind that I have 2 x 24Hz F3 main speakers and 2 x 35Hz F3 surround speakers - so through most of the bass range, I had 5.1 active support speakers

The more your base layer speakers are capable of, the less critical the subs become.

I don't believe any of the other "RoomEQ" solutions assist with decay. (I expect that will change over time, but it may take some years for real competitors to develop...)

 

[middlemarch]

There's a new beta version available on the AVS Dirac Live thread in case anyone hasn't gotten the word, version 3.14.1. Have used it for a bit and it seems stable.

I've also discovered that switching Presets in the receiver while a Dirac session is active makes for trouble when trying to load a new filter set. In my case got a configuration error and a filter getting loaded with the comments field as the file name, followed by the receiver no longer being seen by Dirac. Had to do a power on Restart on the Denon to resolve that one.

My experience so far has been very positive as to results. I've found defaults are pretty darned good. Tried higher support levels but went back to default -18dB. I have raised FSO in some cases, lowered in others, and am still evaluating. I suspect these tweaks only come into play at the extremes, which I haven't explored as yet. I can't play with who supports whom in my system since filters are maxed out at 91 of 92 used. it is what it is.

Biggest finding so far is that my system doesn't like full range EQ. Setting the upper limit to 500Hz for me is much better. I've heard others say that full range "takes the life out of" their speakers, and that's my experience as well. Quite dramatic. I may try even lower than 500Hz.

 

[Oddball]

People are reporting very positive results even from basic 2.0 traditional stereo setups....

With a 5.1.4 setup vs 5.2.4 - the addition of the extra sub made a very very small difference in decay - no substantive subjective difference in performance - but I believe it may have raised my max SPL headroom (a different dimension in terms of parameters)

However keep in mind that I have 2 x 24Hz F3 main speakers and 2 x 35Hz F3 surround speakers - so through most of the bass range, I had 5.1 active support speakers

The more your base layer speakers are capable of, the less critical the subs become.

I don't believe any of the other "RoomEQ" solutions assist with decay. (I expect that will change over time, but it may take some years for real competitors to develop...)

I think we have a bit of divergence of opinions here. Different rooms, different systems and different listening goals all bear to that.

Things need to be put in right perspective and it is probably the desired SPL and distortion associated with it.

Anything can do something at 70dB and without +12dB bass curve. But then if you want reference volume with +12dB curve, another ball game altogether. Your bed channels just fall apart despite the mighty F3 they have. Subs generally don't especially if good sized.

 

[EngineerNate]

I wonder if full range correction sounding good/bad is dependent on how close to the target curve your speakers' natural response is in the higher registers.

 

[mmares056]

There's a new beta version available on the AVS Dirac Live thread in case anyone hasn't gotten the word, version 3.14.1. Have used it for a bit and it seems stable.

I've also discovered that switching Presets in the receiver while a Dirac session is active makes for trouble when trying to load a new filter set. In my case got a configuration error and a filter getting loaded with the comments field as the file name, followed by the receiver no longer being seen by Dirac. Had to do a power on Restart on the Denon to resolve that one.

My experience so far has been very positive as to results. I've found defaults are pretty darned good. Tried higher support levels but went back to default -18dB. I have raised FSO in some cases, lowered in others, and am still evaluating. I suspect these tweaks only come into play at the extremes, which I haven't explored as yet. I can't play with who supports whom in my system since filters are maxed out at 91 of 92 used. it is what it is.

Biggest finding so far is that my system doesn't like full range EQ. Setting the upper limit to 500Hz for me is much better. I've heard others say that full range "takes the life out of" their speakers, and that's my experience as well. Quite dramatic. I may try even lower than 500Hz.

I also used the 3.14.1 version this morning and no issues. Everything worked flawless. I settled on a config where speakers are separated into their own groups including subs with support levels, -18dB for speaker self support and -18dB for speaker support, FSH 150Hz for all speakers. FSL 65Hz for bed layer speakers, and FSL 100Hz for Tops. I found these filter design parameters to sound phenomenal. I do full range EQ and I love how my speakers sound.

 

[mmares056]

I wonder if full range correction sounding good/bad is dependent on how close to the target curve your speakers' natural response is in the higher registers.

I think there is validity to that. For full range correction sounding good, the correct target curve would be a target curve that makes the speaker itself as neutral and accurate as possible in the mid to high frequencies.

 

[middlemarch]

I wonder if full range correction sounding good/bad is dependent on how close to the target curve your speakers' natural response is in the higher registers.

I've been giving this some thought, and perhaps my earlier statement may come off as a bit too "definitive". I think much has to do with your speakers, obviously, and your room and it's characteristics. My speaker front 5 are all Revel with beryllium tweeters, so I kinda like their sound anyway, and my room isn't very treated. Have a thick rug and furniture, but nothing on walls or ceiling. So it could be my particular set of circumstances.

And one factor I realize I should look at more closely. When I run full range my target curve has a 4dB bump in the treble (futile attempt at helping out old ears). What I'm forgetting is I'm sure Dirac drops the average level by that much for headroom. So when I switch between two versions the restricted range version is about 4dB hotter. Of course louder sounds better, so that could be part of what I'm hearing. Not all of it, the difference is pretty dramatic. I have to check that out more closely.

I think the rest is just me liking my natural speaker response even though they don't track the target curve perfectly.

 

[Nivag123]

Hello,

i just saw the new release of Dirac ART version 3.14.1 on 29 Jan 2025. There is a note in the changelog about known issues. The 120Hz low pass filter on the LFE channel introduces a 5ms delay. Is there any additional info on this, what does it mean and what do we do about it?

 

[dlaloum]

There's a new beta version available on the AVS Dirac Live thread in case anyone hasn't gotten the word, version 3.14.1. Have used it for a bit and it seems stable.

I've also discovered that switching Presets in the receiver while a Dirac session is active makes for trouble when trying to load a new filter set. In my case got a configuration error and a filter getting loaded with the comments field as the file name, followed by the receiver no longer being seen by Dirac. Had to do a power on Restart on the Denon to resolve that one.

My experience so far has been very positive as to results. I've found defaults are pretty darned good. Tried higher support levels but went back to default -18dB. I have raised FSO in some cases, lowered in others, and am still evaluating. I suspect these tweaks only come into play at the extremes, which I haven't explored as yet. I can't play with who supports whom in my system since filters are maxed out at 91 of 92 used. it is what it is.

Biggest finding so far is that my system doesn't like full range EQ. Setting the upper limit to 500Hz for me is much better. I've heard others say that full range "takes the life out of" their speakers, and that's my experience as well. Quite dramatic. I may try even lower than 500Hz.

Can I suggest that you also try full range EQ, but using the mains L/C/R frequency response profile as your target curve - that way you let Dirac do its stuff in phase and impulse correction while minimising any EQ adjustments...

Not saying it will definitely be better or worse - but it is a strategy you can use to avoid dulling effects of EQ, while keeping the clarifying effects of phase/impulse correction...

 

[dlaloum]

I think we have a bit of divergence of opinions here. Different rooms, different systems and different listening goals all bear to that.

Things need to be put in right perspective and it is probably the desired SPL and distortion associated with it.

Anything can do something at 70dB and without +12dB bass curve. But then if you want reference volume with +12dB curve, another ball game altogether. Your bed channels just fall apart despite the mighty F3 they have. Subs generally don't especially if good sized.

I still need to do further testing at higher SPL's...

However it is worthy of note that in the bass frequencies, the supported mains have lower THD than in unsupported mode... this, to me, implies that the support speakers have taken over some of the load, and the primary speaker is therefore running at a lower SPL level - and hence a lower distortion level - this would also be expected to provide higher max SPL's... at spec, the speakers should be able to output 109db @ 2m with 256W input (and can handle 350W continuous) - the manual does not specify the level of THD at these high levels however...

I expect the B&O Penta's on my rear channel would be less capable at higher levels (spec says 92db/w, power handling 150W, which implies an easy 110db @ 2m with 128W) - so still ample in theory...

At some point I will get around to trying some measurements at my reference peak level of 95db (as I use 75db as reference)

 

[dlaloum]

I wonder if full range correction sounding good/bad is dependent on how close to the target curve your speakers' natural response is in the higher registers.

That's my thinking too...

 

[EWL5]

Hello,

i just saw the new release of Dirac ART version 3.14.1 on 29 Jan 2025. There is a note in the changelog about known issues. The 120Hz low pass filter on the LFE channel introduces a 5ms delay. Is there any additional info on this, what does it mean and what do we do about it?

It's a known issue so Dirac will fix it when they fix it. There's nothing you can do about it in the meantime.

 

[middlemarch]

Can I suggest that you also try full range EQ, but using the mains L/C/R frequency response profile as your target curve - that way you let Dirac do its stuff in phase and impulse correction while minimising any EQ adjustments...

Not saying it will definitely be better or worse - but it is a strategy you can use to avoid dulling effects of EQ, while keeping the clarifying effects of phase/impulse correction...

Thanks for that suggestion, I will give it a go, in addition to resolving the treble bump question. I want to make clear that full range doesn't sound "bad" to me, if it wasn't for the A/B comparison I'd be very happy with it.

Best way I can describe it is analogous to Amir's subjective impressions when he tested the JBL 4329. It wasn't as accurate as his reference system but he very much enjoyed the sound.

 

[tifune]

It's a known issue so Dirac will fix it when they fix it. There's nothing you can do about it in the meantime.

Well, not with that attitude! Seems like a fair Q. Does it mean we should delay all other other channels by 5ms? Or is Dirac already compensating for us? The release notes aren't clear on what, if anything, we should do about it

 

[PGAMiami]

But does it mean we should delay all other other channels by 5ms? Or is Dirac already compensating for us? The release notes aren't clear on what, if anything, we should do about it

If all we’re talking about is the LFE then it should not affect speach, music or most sounds. I’m much less worried about getting that explosion just right than getting music right.